论文部分内容阅读
高分辨率影像为矿产资源开发遥感监管提供了更为精确有效的数据。以霍林河露天煤矿区为研究区,应用高分一号卫星影像为主要数据源,在面向对象的影像分类基础上,探讨了露天煤矿区用地类型信息提取优先顺序对最终分类精度的影响。结果表明:露天煤矿区的用地类型信息提取中,采用优先提取采矿场和排土场等资源开发用地类型、而后提取其他非开发用地的优先级顺序的分类精度最高,其总体精度达到82%,Kappa系数达到0.78,可以为露天煤矿区的用地类型信息提取提供理论和方法支持。
High-resolution images provide more accurate and efficient data for remote sensing monitoring of mineral resources development. Taking Huolinhe opencast coal mine as the research area and the high score of No.1 satellite imagery as the main data source, this paper discusses the influence of the prioritized order of the information of the landform type on the final classification accuracy based on object-oriented image classification. The results show that in the extraction of land use type information in opencast coal mining areas, the priority is to extract the types of land for resource exploitation such as mining and dumping sites, and then the priority order of extracting other non-development land is the highest. The overall accuracy is 82% Kappa coefficient of 0.78, which can provide theory and method support for the extraction of land type information in open coal mine.